[developers] PET question

[Bart Cramer]

Hi all,

a small addendum: the rather large commit in revision 498 seems to be
the culprit. 497 still works fine for me.

opt_mrs
cheap: ../common/configs-inner.h:254: Config::option<T>&
Config::get_option_checked(const std::string&) [with T = const char*]:
Assertion `false' failed.
Error in format: No more arguments.
  Serious signal ~D caught.
                  ^
while processing indirect format string:
  ~?
   ^
Broken at NIL.
MT>

HTH,

Bart.

On Sun, Nov 23, 2008 at 1:36 PM, Bart Cramer <bart.cramer at gmail.com> wrote:
> Dear all,
>
> Since a couple of revisions, I am running into problems when I want to
> export trees to tsdb. I am using this command line:
>
> cat /home/HU/bcramer/Tiger2HPSG/data/tiger.1000.txt |
> /home/HU/bcramer/delphin/pet/cheap/cheap
> /home/HU/bcramer/Tiger2HPSG/grammar/german.grm -packing=15
> -nsolutions=1000 -limit=50000 -tsdbdump
> /home/HU/bcramer/tsdb/Tiger-first-1K/
>
> However, this is what I get:
>
> [bcramer at cluster-04 Tiger2HPSG]$ ./parse-dev
> Error in format: No more arguments.
>  Serious signal ~D caught.
>                  ^
> while processing indirect format string:
>  ~?
>   ^
> Broken at NIL.
> MT>
>
> Removing the -tsdbdump option results in expected behaviour.
>
> Two revisions earlier, it seemed to have something to do with the
> option opt_mrs, but now it's not mentioned anymore. Also, previously
> it only happened when the tsdb dump was started, i.e. after the first
> sentence has been parsed, but now it does it before.
>
> Is this something that other people experience, too?
>
> Bart.
>